Get in Touch

Course Outline

Introduction

  • What is Kubefirst?
  • Overview of Kubefirst features and architecture
  • Integration with open-source tools (Terraform, Vault, Argo, etc.)

Setting up the Development Environment

  • Installing and configuring the Kubefirst SDK.
  • Selecting and setting up cloud platforms (Local, AWS, Civo, etc.).
  • Creating a Kubefirst project in Visual Studio Code.
  • Exploring the project structure and files.

Kubefirst Programming Model

  • Understanding Kubefirst directives and clauses.
  • Using parallel directives to create parallel regions.
  • Using kernels directives to parallelize kernels.
  • Using loop directives to parallelize loops.

Kubefirst Data Management

  • Managing data movement using data directives.
  • Synchronizing data with update directives.
  • Enhancing data reuse with cache directives.
  • Creating device functions with routine directives.

Kubefirst API and Libraries

  • Understanding the Kubefirst API within host programs.
  • Using the Kubefirst API to query device information.
  • Setting device number and type using the Kubefirst API.
  • Handling errors and exceptions with the Kubefirst API.
  • Utilizing Kubefirst libraries (math, random, complex).

Kubefirst Tools

  • Understanding the role of Kubefirst tools.
  • Using tools for profiling and debugging Kubefirst programs.
  • Measuring and improving execution time with tools (PGI Compiler, NVIDIA Nsight Systems, Allinea Forge).

Kubefirst Applications

  • Best practices for developing and deploying with Kubefirst.
  • Interoperating with other programming models (CUDA, OpenCL, SYCL).
  • Integrating with GPU libraries (cuBLAS, cuFFT).
  • Demonstrating the Kubefirst Metaphor demo application.

Summary and Next Steps

Requirements

  • Basic understanding of computer architecture and memory hierarchy.
  • Familiarity with Docker and Kubernetes concepts and commands.

Target Audience

  • Developers.
 14 Hours

Testimonials (3)

Related Categories